When there is enough anecdotal evidence, that is consistent, you have to consider that the researcher is missing something. In this case, there might just be a

When there is enough anecdotal evidence, that is consistent, you have to consider that the researcher is missing something. In this case, there might just be a misunderstanding - the economist is considering all jobs, whereas this evidence is coming from experienced tech jobs. The evidence is multi-faceted - people quitting, more hits from recruiters, etc.